The approaches described in this section are approaches that could be pursued, but not necessarily approaches that have been previously conceived or pursued. Therefore, unless otherwise indicated, it should not be assumed that any of the approaches described in this section qualify as prior art merely by virtue of their inclusion in this section.
Satellite imagery is commonly used in computer-implemented digital map programs and applications. Typically, digital images that are received at a computer system from a satellite have to be adjusted and processed into a format that map applications can use. For example, satellite images may have to be combined together and divided into smaller image tiles that correspond to area of a digital map.
Satellite images are taken at a high resolution and at large quantities, which requires significant amounts of storage. A solution for reducing storage requirements is to store the images in a lossy compressed format. However, converting the satellite images to a lossy compressed format results in image artifacts within the converted digital image. The artifacts may result in errors, such as noticeable seams or artifacts along overlapping image areas, when the satellite images are combined. Combining images can result, for example, when images representing adjacent tiles of a map are joined to form a view of a multi-tile area in a map display application. When lossy compression is used to store the tiles, combining tiles can result in jagged edges, black lines, or other visually undesirable errors in the combined image at the joint location. Therefore, there is a need for a technique for reducing or eliminating the appearance of artifacts from digital images to allow for more seamless combined images.